Evals for Every Language - Language bew — leaderboard
Metric: Average Score (%). Source: huggingface.co. 68 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 100 |
| 2 | Claude Opus 4.7 | 100 |
| 3 | GPT-5.4 | 100 |
| 4 | Claude Opus 4.8 | 100 |
| 5 | DeepSeek V3.1 | 100 |
| 6 | GLM 4.5 Air | 100 |
| 7 | nova-2-lite-v1 | 100 |
| 8 | Gemini 3.1 Pro (Preview) | 96.67 |
| 9 | GPT-5 | 96.67 |
| 10 | GPT-5.2 | 96.67 |
| 11 | Claude Sonnet 4.5 | 96.67 |
| 12 | GPT-5.5 | 96.67 |
| 13 | Qwen 3 30B A3B 2507 Instruct | 96.67 |
| 14 | Ministral 3 8B | 96.67 |
| 15 | Mistral Small 3.2 | 96.67 |
Interactive version: theaggregate.ai/benchmark?slug=evals-for-every-language-language-bew · How the rankings work · Data refreshed daily, snapshot 2026-07-22.